Using mathematical morphology, recent research has clarified the geometrical properties of topological transformation procedures such as skeletonization and boundary detection, which are typical operations for shape analysis. Further, mathematical morphology has derived what we must compute parallelly for shape analysis. Mathematical morphology, however, does not provide how we must compute parallelly each operation because the theory is based on integral geometry and is not based on the theory of parallel computation. A well-defined language clarifies how algorithms should be compute parallelly.1 Then, it is necessary to introduce a language to describe the parallelism of morphological operations. The theory of computation2 is unfamiliar to image analysis society. Then, just as mathematical morphology brought new methodology to image analysis society, the computational theory will enrich shape analysis.
KeywordsShape Analysis Mathematical Morphology Morphological Operation Integral Geometry Digital Picture
Unable to display preview. Download preview PDF.
- 4.A. Imiya and T. Nakamura, “Morphological operation via convolution”, Proc. of 6th Scandinavian Conference on Image Analysis, vol. 2, pp. 878–881, Oulu, Finland, June 1989.Google Scholar
- 5.J.-I. Toriwaki, Digital Image Processing for Image Understanding, (in Japanese), vol.1 and vol.2, Tokyo: Syokodo, 1988.Google Scholar
- 6.M. Hušek, “Categories and mathematical morphology”, in Categorical methods in computer science with aspects from topology, H. Ehring et al. Eds. Lecture Notes in Computer Science, vol. 393, Belrin: Springer-Verlag, 1989, pp. 295–301.Google Scholar